Despite evidence of pollinator declines from many regions across the globe, the threat this poses to plant populations is not clear because plants can often produce seeds without animal pollinators. Here, we quantify pollinator contribution to seed production by comparing fertility in the presence versus the absence of pollinators for a global dataset of 1174 plant species. We estimate that, without pollinators, a third of flowering plant species would produce no seeds and half would suffer an 80% or more reduction in fertility. Pollinator contribution to plant reproduction is higher in plants with tree growth form, multiple reproductive episodes, more specialized pollination systems, and tropical distributions, making these groups especially vulnerable to reduced service from pollinators. These results suggest that, without mitigating efforts, pollinator declines have the potential to reduce reproduction for most plant species, increasing the risk of population declines.

This novel approach, which demonstrates non-random assembly of complex flower colour patterns and corroborates their functional association with particular pollinators, provides strong evidence that pollinators influence plant community assembly. Results suggest that in some community contexts the benefits of pollinator sharing outweigh the costs of heterospecific pollen transfer, generating clustered assembly. They also challenge the perception of generalized pollination in daisies, suggesting instead that complex daisy colour patterns represent a pollination syndrome trait linked to specific fly pollinators.

Aim Insect distribution patterns can result from historical contingency (biogeography and dispersal limitation), abiotic filtering and biotic factors (ecological interactions and evolutionary associations). Here, we analyse turnover of plant and insect herbivore community composition at multiple spatial scales to tease apart these influences. While positive associations between plant and insect turnover across broad spatial scales could arise through any of these influences, strong association at very local scales is only likely if insect distributions are determined primarily by biotic factors (i.e. host specificity).Location The Cape Floristic Region (CFR), South Africa.Methods To characterize the relationship between spatial turnover in plant and insect composition in the CFR, communities of Restionaceae, a dominant family in the florally diverse CFR, and their associated herbivores were sampled using a spatially nested sampling design on three spatially separated mountain blocks with similar climates, thus controlling for broad abiotic influences. This allowed us to quantify insect and plant turnover, and their association, at multiple independent spatial scales. Redundancy analysis was used to determine the effects of plant on insect composition, controlling for geographical distance.Results Insect species turnover was significantly related to plant species and phylogenetic turnover at local, as well as broad, spatial scales, suggesting that insect distribution patterns are mainly structured by host specificity. Plant communities show near complete turnover at small spatial scales (i.e. communities situated 0.1-3 km apart), with insects mirroring this pattern. Further, insect turnover increased significantly with increasing geographical separation (e.g. between mountains), suggesting an additional influence of biogeographical factors on insect distributions in the CFR. Measured environmental and plant structural components had no influence on insect composition.Main conclusions High insect beta diversity positively associated with plant turnover at local scales suggests insect herbivore diversity patterns in the CFR are primarily structured by plant distribution patterns.

Recent work has suggested that emergent ecological network structure exhibits very little spatial or temporal variance despite changes in community composition. However, the changes in network interactions associated with turnover in community composition have seldom been assessed. Here we examine whether changes in ecological networks are best detected by standard emergent network metrics or by assessing internal network changes (i.e. interaction and composition turnover). To eliminate possible spatial or phylogenetic effects, that in large‐scale studies may obscure mechanisms structuring networks and interactions, we sampled multiple antagonistic (plant–herbivore) networks for a single diverse plant family (the Restionaceae) in the hyperdiverse Cape Floristic Region. These are the first plant–herbivore networks constructed for this global biodiversity hotspot. We found invariant emergent network structure despite considerable changes in insect and plant composition across communities over time and space. In contrast, there was high interaction turnover between networks. Seasonally, this was driven by turnover in insect species and insect host switching. Spatially, this was driven by simultaneous turnover in plant and insect species, suggesting that many insects are host specific or that both groups exhibit parallel responses to environmental gradients. Spatial interaction turnover was also driven by turnover in plants, showing that many insects can utilise multiple (possibly closely related) hosts and this may create divergent selection gradients that promote insect speciation. Thus we show highly variable interaction fidelity, despite invariant emergent network structure. We suggest that evaluating internal network changes may be more effective at elucidating the processes structuring networks, and many fine‐scale changes may be obscured when only calculating emergent network metrics.

Globally plant species richness is a significant predictor of insect richness. Whether this is the result of insect diversity responding directly to plant diversity, or both groups responding in similar ways to extrinsic factors, has been much debated. Here we assess this relationship in the Cape Floristic Region (CFR), a biodiversity hotspot. The CFR has higher plant diversity than expected from latitude (i.e., abiotic conditions), but very little is known about the diversity of insects residing in this region. We first quantify diversity relationships at multiple spatial scales for one of the dominant plant families in the CFR, the Restionaceae, and its associated insect herbivore community. Plant and insect diversity are significantly positively correlated at the local scales (10–50 m; 0.1–3 km), but not at the regional scales (15–20 km; 50–70 km). The local scale relationship remains significantly positively correlated even when accounting for the influence of extrinsic variables and other vegetation attributes. This suggests that the diversity of local insect assemblages may be more strongly influenced by plant species richness than by abiotic variables. Further, vegetation age and plant structural complexity also influenced insect richness. The ratio of insect species per plant species in the CFR is comparable to other temperate regions around the world, suggesting that the insect diversity of the CFR is high relative to other areas of the globe with similar abiotic conditions, primarily as a result of the unusually high plant diversity in the region.

Premise Plants can mitigate the fitness costs associated with pollen consumption by floral visitors by optimizing pollen release rates. In buzz‐pollinated plants, bees apply vibrations to remove pollen from anthers with small pores. These poricidal anthers potentially function as mechanism staggering pollen release, but this has rarely been tested across plant species differing in anther morphology. Methods In Solanum Section Androceras, three pairs of buzz‐pollinated species have undergone independent evolutionary shifts between large‐ and small‐flowers, which are accompanied by replicate changes in anther morphology. We used these shifts in anther morphology to characterize the association between anther morphology and pollen dispensing schedules. We applied simulated bee‐like vibrations to anthers to elicit pollen release, and compared pollen dispensing schedules across anther morphologies. We also investigated how vibration velocity affects pollen release. Results Replicate transitions in Solanum anther morphology are associated with consistent changes in pollen dispensing schedules. We found that small‐flowered taxa release their pollen at higher rates than their large‐flowered counterparts. Higher vibration velocities resulted in quicker pollen dispensing and more total pollen released. Finally, both the pollen dispensing rate and the amount of pollen released in the first vibration were negatively related to anther wall area, but we did not observe any association between pore size and pollen dispensing. Conclusions Our results provide the first empirical demonstration that the pollen dispensing properties of poricidal anthers depend on both floral characteristics and bee vibration properties. Morphological modification of anthers could thus provide a mechanism to exploit different pollination environments.

Floral apparency is shaped by both mutualistic and antagonistic interactions that can act in opposing ways. Pollinators are expected to select for more visually apparent flowers, but this likely trades off against the potentially severe fitness costs of damage to apparent flowers by floral herbivores. One way in which flowers that close during parts of the day might circumvent this trade‐off is by evolving less visible lower petal surfaces that are inconspicuous to herbivores when flowers are closed. Here, we used visual system modelling and herbivory experiments to test whether petal surfaces that are exposed when flowers are closed are cryptically coloured. We collected lower and upper petal surface spectra for 77 Asteraceae species from Namaqualand, South Africa. This included closing species that expose their lower petal surfaces for 5–6 daylight hours and non‐closing species that do not expose their lower surfaces. We used these contrasting groups to test the expectation of reduced conspicuousness of lower petal surfaces in closing, but not non‐closing species. By modelling reflectance spectra of petal surfaces against a green leaf background in various visual systems, we showed (a) that conspicuousness of upper petal surfaces to pollinators and various herbivores was strongly correlated, suggesting the potential for fitness trade‐offs between attracting mutualists and antagonists to open flowers, (b) that closing species' lower petal surfaces were less visible to herbivores against a leaf background than those of non‐closing species and (c) that closing species had larger differences between upper and lower petal surface coloration than non‐closing species. Behavioural experiments with tortoise herbivores demonstrated that flowers are easily detected when upper surfaces are exposed, but that tortoises were unable to distinguish lower petal surfaces against a leaf background, resulting in reduced flower herbivory. These results are consistent with selection by herbivores for cryptic coloration of lower petal surfaces, and divergence of coloration between lower and upper petal surfaces in species with closing flowers. Visual crypsis of flowers may be an effective anti‐herbivory strategy during times when pollinators are inactive, and provides an alternative to chemical defence, which often involves costs to pollination. A free Plain Language Summary can be found within the Supporting Information of this article.

The striking variation in flower color across and within Angiosperm species is often attributed to divergent selection resulting from geographic mosaics of pollinators with different color preferences. Despite the importance of pollinator mosaics in driving floral divergence, the distributions of pollinators and their color preferences are seldom quantified. The extensive mass-flowering displays of annual daisy species in Namaqualand, South Africa, are characterized by striking color convergence within communities, but also color turnover within species and genera across large geographic scales. We aimed to determine whether shifts between orange and white-flowered daisy communities are driven by the innate color preferences of different pollinators or by soil color, which can potentially affect the detectability of different colored flowers. Different bee-fly pollinators dominated in both community types so that largely non-overlapping pollinator distributions were strongly associated with different flower colors. Visual modeling demonstrated that orange and white-flowered species are distinguishable in fly vision, and choice experiments demonstrated strongly divergent color preferences. We found that the dominant pollinator in orange communities has a strong spontaneous preference for orange flowers, which was not altered by conditioning. Similarly, the dominant pollinator in white communities exhibited an innate preference for white flowers. Although detectability of white flowers varied across soil types, background contrast did not alter color preferences. These findings demonstrate that landscape-level flower color turnover across Namaqua daisy communities is likely shaped by a strong qualitative geographic mosaic of bee-fly pollinators with divergent color preferences. This is an unexpected result given the classically generalist pollination phenotype of daisies. However, because of the dominance of single fly pollinator species within communities, and the virtual absence of bees as pollinators, we suggest that Namaqua daisies function as pollination specialists despite their generalist phenotypes, thus facilitating differentiation of flower color by pollinator shifts across the fly pollinator mosaic.